The transcript of the AEA session that Brad Delong moderated is worth reading.
Let me mention something on which I have shifted my views. Johannes Wieland is from Berkeley. In his job market paper he says that in new Keynesian models at the zero lower bound there is, theoretically, a perverse reaction to supply shocks. An adverse supply shock which reduces potential output rises expected inflation and so reduces real interest rates and boosts output. this topsy-turvy effect enables an adverse supply shock to actually increase output at the zero lower bound. He goes out and tries to see if this is actually true. He finds the opposite is true: if you raise oil prices it is actually bad for Japanese output and not good at the zero lower bound. The world seems to have a more conventional reaction. He comes to the conclusion that this tells us something about the model—if you have a model in which everything is determined by sticky prices, in crisis times you can see that a better fit is attained by models that say that you really ought to look at balance sheets. Other research has recently emphasized balance sheets as very important in what has been happening in the past several years. There are lots and lots of open questions out there that I wish we just understood more.